{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.9242372300308537, "global_step": 4264, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.22, "learning_rate": 6.115909912286932e-07, "loss": 1.0478, "step": 328 }, { "epoch": 0.22, "eval_loss": 0.20066721737384796, "eval_runtime": 421.2422, "eval_samples_per_second": 7.122, "eval_steps_per_second": 0.89, "eval_wer": 0.2669276913682244, "step": 328 }, { "epoch": 0.45, "learning_rate": 9.100780032860387e-07, "loss": 1.0471, "step": 656 }, { "epoch": 0.45, "eval_loss": 0.2007066309452057, "eval_runtime": 420.7757, "eval_samples_per_second": 7.13, "eval_steps_per_second": 0.891, "eval_wer": 0.26699953302920365, "step": 656 }, { "epoch": 0.67, "learning_rate": 8.2992075702454e-07, "loss": 1.031, "step": 984 }, { "epoch": 0.67, "eval_loss": 0.20067130029201508, "eval_runtime": 419.7329, "eval_samples_per_second": 7.147, "eval_steps_per_second": 0.893, "eval_wer": 0.266245195588922, "step": 984 }, { "epoch": 0.9, "learning_rate": 7.497635107630415e-07, "loss": 1.044, "step": 1312 }, { "epoch": 0.9, "eval_loss": 0.20047353208065033, "eval_runtime": 419.6669, "eval_samples_per_second": 7.149, "eval_steps_per_second": 0.894, "eval_wer": 0.265885987284026, "step": 1312 }, { "epoch": 1.12, "learning_rate": 6.698506463499012e-07, "loss": 1.0448, "step": 1640 }, { "epoch": 1.12, "eval_loss": 0.2001875340938568, "eval_runtime": 423.812, "eval_samples_per_second": 7.079, "eval_steps_per_second": 0.885, "eval_wer": 0.2658141456230468, "step": 1640 }, { "epoch": 1.35, "learning_rate": 5.896934000884026e-07, "loss": 1.0489, "step": 1968 }, { "epoch": 1.35, "eval_loss": 0.2000104784965515, "eval_runtime": 432.0862, "eval_samples_per_second": 6.943, "eval_steps_per_second": 0.868, "eval_wer": 0.2658500664535364, "step": 1968 }, { "epoch": 1.57, "learning_rate": 5.09536153826904e-07, "loss": 1.0277, "step": 2296 }, { "epoch": 1.57, "eval_loss": 0.200164794921875, "eval_runtime": 428.8208, "eval_samples_per_second": 6.996, "eval_steps_per_second": 0.874, "eval_wer": 0.26552677897913, "step": 2296 }, { "epoch": 1.8, "learning_rate": 4.2937890756540545e-07, "loss": 1.0364, "step": 2624 }, { "epoch": 1.8, "eval_loss": 0.2002057284116745, "eval_runtime": 426.0723, "eval_samples_per_second": 7.041, "eval_steps_per_second": 0.88, "eval_wer": 0.2659578289450052, "step": 2624 }, { "epoch": 2.02, "learning_rate": 3.492216613039069e-07, "loss": 1.035, "step": 2952 }, { "epoch": 2.02, "eval_loss": 0.20034705102443695, "eval_runtime": 425.252, "eval_samples_per_second": 7.055, "eval_steps_per_second": 0.882, "eval_wer": 0.26635295808039083, "step": 2952 }, { "epoch": 2.25, "learning_rate": 2.690644150424083e-07, "loss": 1.0306, "step": 3280 }, { "epoch": 2.25, "eval_loss": 0.20014077425003052, "eval_runtime": 418.7166, "eval_samples_per_second": 7.165, "eval_steps_per_second": 0.896, "eval_wer": 0.2662092747584324, "step": 3280 }, { "epoch": 2.47, "learning_rate": 1.8890716878090973e-07, "loss": 1.0395, "step": 3608 }, { "epoch": 2.47, "eval_loss": 0.2000894397497177, "eval_runtime": 418.0797, "eval_samples_per_second": 7.176, "eval_steps_per_second": 0.897, "eval_wer": 0.2654908581486404, "step": 3608 }, { "epoch": 2.7, "learning_rate": 1.0874992251941118e-07, "loss": 1.0426, "step": 3936 }, { "epoch": 2.7, "eval_loss": 0.20004504919052124, "eval_runtime": 416.8848, "eval_samples_per_second": 7.196, "eval_steps_per_second": 0.9, "eval_wer": 0.2658141456230468, "step": 3936 }, { "epoch": 2.92, "learning_rate": 2.883705810627083e-08, "loss": 1.0431, "step": 4264 }, { "epoch": 2.92, "eval_loss": 0.20020076632499695, "eval_runtime": 419.7762, "eval_samples_per_second": 7.147, "eval_steps_per_second": 0.893, "eval_wer": 0.2656345414705988, "step": 4264 } ], "max_steps": 4374, "num_train_epochs": 3, "total_flos": 1.8745213924829307e+19, "trial_name": null, "trial_params": null }